On Tue, 25 Oct 2022 12:59:20 -0400, Al Loeffler wrote: >Here's a USERMOD to set the block size to 0 for z/OS 2.5. I have a USERMOD for >each release going back to 1.11. > That seems dangerous. A future release might change the offsets (which "++VER" will probably detedt.) And it's likely to fail with BSAM using hard-coded buffer sizes.
*** An RFE would be far better. *** IBM might be averse to such a "performance" RFE because it might set a precedent for RFEs to numerous antiquated products. >++USERMOD(XTEMXMI) > /* > Transmit OUTDA() forces the BLKSIZE of the output data set > to 3120 (X'0C30'). > > This USERMOD sets it to 0 (system determined block size). > > FMID(HTE77D0) TSO/E 2.5 > */ . >++VER(Z038) FMID(HTE77D0). >++ZAP (INMXM). > NAME INMXM > VER 1A7C 0000,0C30 > REP 1A7C 0000,0000 > >Al Loeffler -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
